Although a high proportion of patients seen in Emergency Departments (EDs) have at-risk or problem alcohol use, few are screened and receive services such as brief interventions (BI) designed to help them cut-back or stop drinking. EDs do not routinely provide BIs, perhaps due to feasibility challenges such as training of staff, monitoring fidelity, and maintaining a system to ensure longer-term implementation. Alcohol BIs have been found to be efficacious and effective in a variety of health care settings. However, the evidence for their use in the ED has been mixed. There is a pressing need to develop efficacious strategies to screen and optimally deliver alcohol BIs in this fast-paced and widely-used setting. Existing clinician-delivered BI strategies need to be modified so that they can be standardized and administered with high fidelity and minimal demands on ED staff time and resources. Computer-delivered BIs are one method to address the challenges inherent in delivering interventions in this and other healthcare settings. The proposed study will use computerized screening via touch-screen computer tablets with audio to recruit 750 inner-city ED patients screening positive for at-risk or problem alcohol use. Participants age 18-60 will be randomized to one of three conditions: 1) Computer-delivered brief intervention (C-BI; n=250); 2) Therapist-delivered brief intervention (T-BI; n=250); or 3) Enhanced usual care (EUC; n=250). All participants will receive written information regarding community resources; individuals who meet alcohol abuse/dependence criteria will also receive alcohol treatment referrals. Stratified random assignment [by gender; meeting criteria for an alcohol use disorder - yes/no] will take place at baseline for all conditions. The aims of the study are to develop and refine tailored motivational brief interventions that are parallel in structure but have varied delivery modalities (computer vs. therapist) for patients with at-risk or problematic alcohol use, and to conduct a randomized controlled trial comparing the efficacy of these BI approaches (C-BI, T-BI, control) on subsequent alcohol consumption and alcohol consequences, including alcohol-related injury, mental and physical-health functioning, and HIV risk behaviors at 3-, 6-, and 12-months post-ED visit. The rigorous examination of the efficacy of therapist- vs. computer- delivered BIs, including potential moderators and mediators, will address the key limitations raised by previous trials and will determine the optimal modality for wide implementation of brief alcohol interventions in this venue. Because the ED is such an important portal for entry into the medical care system, particularly for inner-city patients, the delivery of efficacious alcohol BIs that emphasize key motivational interviewing components and minimize staff resources could have a major public health impact..